Skip to content

Actions: lucidrains/native-sparse-attention-pytorch

Actions

Pytest

Actions

Loading...
Loading

Show workflow options

Create status badge

Loading
test.yml
98 workflow run results
98 workflow run results

Filter by Event

Filter by Status

Filter by Branch

Filter by Actor

more cleanup Pytest #99: Commit 059bf78 pushed by lucidrains
1m 50s main
main
1m 50s
cleanup triton flash attn Pytest #98: Commit 2f14297 pushed by lucidrains
1m 41s main
main
1m 41s
copy paste in a working impl of flash attention from ring-attention-p… Pytest #97: Commit b1dee31 pushed by lucidrains
1m 47s main
main
1m 47s
redo with approach of using compressed similarities for interpolation… Pytest #96: Commit 582b844 pushed by lucidrains
1m 41s 0.0.47
0.0.47
1m 41s
redo with approach of using compressed similarities for interpolation… Pytest #95: Commit 582b844 pushed by lucidrains
1m 46s main
main
1m 46s
mask out the block diagonal for the importance score, as block causal… Pytest #94: Commit 16d9bbc pushed by lucidrains
1m 46s 0.0.45
0.0.45
1m 46s
mask out the block diagonal for the importance score, as block causal… Pytest #93: Commit 16d9bbc pushed by lucidrains
1m 44s main
main
1m 44s
when doing interpolation of importance score, remask to 0 for illegal… Pytest #92: Commit 36837b2 pushed by lucidrains
1m 45s 0.0.44
0.0.44
1m 45s
when doing interpolation of importance score, remask to 0 for illegal… Pytest #91: Commit 36837b2 pushed by lucidrains
1m 58s main
main
1m 58s
default to one mem kv for compressed attn Pytest #90: Commit 463963b pushed by lucidrains
1m 44s 0.0.43
0.0.43
1m 44s
default to one mem kv for compressed attn Pytest #89: Commit 463963b pushed by lucidrains
1m 38s main
main
1m 38s
fixes Pytest #88: Commit dfd3b15 pushed by lucidrains
1m 41s main
main
1m 41s
allow for ablating fine block selection for negative control Pytest #87: Commit 0ad8c5e pushed by lucidrains
1m 43s 0.0.42
0.0.42
1m 43s
allow for ablating fine block selection for negative control Pytest #86: Commit 0ad8c5e pushed by lucidrains
1m 46s main
main
1m 46s
ready to be compared with full attention. Pytest #85: Commit e9476ec pushed by lucidrains
1m 40s 0.0.41
0.0.41
1m 40s
ready to be compared with full attention. Pytest #84: Commit e9476ec pushed by lucidrains
1m 36s main
main
1m 36s
oops Pytest #83: Commit f32ed38 pushed by lucidrains
1m 43s 0.0.40
0.0.40
1m 43s
oops Pytest #82: Commit f32ed38 pushed by lucidrains
1m 35s main
main
1m 35s
do the differential topk gating in a more suboptimal way, but accommo… Pytest #81: Commit cfa6e08 pushed by lucidrains
1m 39s 0.0.39
0.0.39
1m 39s
do the differential topk gating in a more suboptimal way, but accommo… Pytest #80: Commit cfa6e08 pushed by lucidrains
1m 50s main
main
1m 50s
setup for students Pytest #79: Commit 671534a pushed by lucidrains
1m 40s main
main
1m 40s
account for learned memory key values in flex compress mask, also cle… Pytest #78: Commit e40b35f pushed by lucidrains
1m 47s 0.0.38
0.0.38
1m 47s
account for learned memory key values in flex compress mask, also cle… Pytest #77: Commit e40b35f pushed by lucidrains
1m 40s main
main
1m 40s
account for learned memory key values in flex compress mask, also cle… Pytest #76: Commit bc31eb7 pushed by lucidrains
1m 55s 0.0.37
0.0.37
1m 55s
account for learned memory key values in flex compress mask, also cle… Pytest #75: Commit bc31eb7 pushed by lucidrains
2m 6s main
main
2m 6s